Senior Project Report PV Hybrid Inverter & BESS

Home , Photovoltaics, Solar energy

Electrical Engineering Department California Polytechnic State University

Senior Project Report

PV Hybrid Inverter & BESS June 14th, 2019

William Dresser Owen McKenzie Derek Seaman Jacob Sussex Jonathan Wharton

Professor William Ahlgren & Professor Ali Dehghan Banadaki

Table of Contents

Table of Contents 1

List of Tables 2

List of Figures 2

Abstract 3

1. General Introduction and Background: 4

2. Overview of Customer Need: 6

3. Project Description: 6

4. Market Research: 7

5. Customer Archetype: 8

6. Market Description: 9

7. Business Model Canvas Graphic: 12

8. Marketing Requirements: 13

9. Block Diagram: 16 10. Requirements: 16 11. Cost Analysis: 17 12. Team Coordination: 18 13. Design Iterations and SolidWorks 18 14.Assembly Iteration and Challenges 26

15. Project Significance and Applications 28 16. Peripherals 29

17. Project Continuation 31

18. Department Future 33

19. References: 34

Appendix A: Senior Project Analysis 36

Appendix B: Preliminary Design Analysis 45

List of Tables

Table 1 Customer Archetype

Table 2 Number of Panels in Configuration and Relevant Metrics

Table 3 Requirements

Table 4 Bill of Materials

Table 5 Budget for Unistrut Cart Design

Table 6 Budget for 8020 Cart Design

Table 7 Budget for Electrical Components

List of Figures

Figure 1 Photovoltaic Cell in a Solar Panel

Figure 2 Cost Trend of Solar Power

Figure 3 Diagram of Tabuchi EIBS16GU2 Energy Flow

Figure 4 Business Model Canvas

Figure 5 Graphical Representation of the Number of Panels in PV Array vs Charge Time

Figure 6 Block Diagram of How the Tabuchi System will be Implemented

Figure 7 First Iteration of 8020 Cart in SolidWorks

Figure 8 2D CAD View of First 8020 Cart Design

Figure 9 2nd Iteration of Cart with Unistrut

Figure 10 Final Cart Design

Figure 11 Phase Converter to be Purchased in Future

Abstract The storage of energy from renewable sources such as photovoltaic based systems is a growing market, with 36 MWh of storage installed in Q1 of 2018. A report from EnergySage earlier this year states that in 2017, 74% of residential solar owners were also interested in energy storage systems. Mainstream systems like Tesla’s Powerwall are competing with other lithium-ion based storage systems from a wide number of providers on the market today. Short term and long term data collection on a system like this could be useful in designing future systems which perform better than the current market offerings.

This project seeks to install and operate the Tabuchi EIBS that the Cal Poly Electrical

Engineering department currently owns. EIBS stands for Eco Intelligent Battery System, and it is meant to be used in conjunction with a photovoltaic array in a residence. This kind of system is a parallel to a source like a Tesla Powerwall, and uses two 10 kWh Li-Ion batteries. As of right now the system is being rolled between room 102 and room 146 of building 20, where it is assembled and ready to be energized. This team would like to assemble, mount, and integrate this system into the EE building micro-grid allowing for future students to test the storage and economical benefit of this system while connected to either the grid or a photovoltaic array. Our first priority was mounting the system on a mobile platform to enable flexible usage wherever its power would be most beneficial. After installation, time permitting, we wish to measure characteristics such as battery life, charge time, switching time, maximum throughput and how efficiently the batteries charge and discharge the energy to be stored.

1. General Introduction and Background:

Solar power is becoming more and more popular due to rising efficiencies and cost effectiveness. In fact, renewable energy generation grew by 167 GW in 2017, representing a stable growth rate of 8.3% according to the SDG Knowledge Hub, a site that documents sustainable development [8].

Solar panels are large panels that are made of many smaller units called photovoltaic cells. These cells use the photons irradiated by the sun to free electrons from the atoms in the photovoltaic cell which generates a current that can be used to charge batteries, power a house, or any other application that requires reasonable amounts of electrical energy.

Figure 1: Photovoltaic Cells in a Solar Panel

When solar panels first came out around 1977 the price for the electricity they produced was about $76.67 per watt, but as of 2013 solar costs about $0.74 per watt (according to Clean

Technica) which makes solar a much better investment today.

Figure 2: Cost Trend of Solar Power

To really get the most out of a photovoltaic system, an energy storage system can be used to capture the energy that is generated by the photovoltaics. By doing this, the energy can then be stored and either sold back to the electric company if it is not used, or the energy can be used during peak hours and be replenished the following day. The benefit of using the energy during peak periods of electricity usage is that the customer can typically pay less for electricity than they normally would be if they were taking power from the grid. In both cases the resident or business saves money compared to the usual practice of relying entirely on the grid for power consumption.

Our project is a system that integrates both photovoltaic and grid-based power. The product will have an inverter and high-capacity batteries mounted on a mobile platform so that the energy can be used in different locations, or to power systems that cannot be turned off. EIBS in general will allow customers to save money by running off of battery power during peak electricity hours, but in our case we hope to use the system primarily as an education and research tool.

2. Overview of Customer Need:

With the usage of renewable energy becoming more and more prevalent, a method for storing excess power generated via renewable sources is an obvious next step. Without a storage system, customers with solar panels are forced to rely on grid power during the evening, or when weather restricts daylight hours. A storage platform enables the user to save money and stay self-reliant under a variety of adverse conditions. In addition, our system focuses on portability, with the inverter and batteries able to be transported separately and easily for testing in multiple locations. This allows users the ability to transport a power source directly to the electronics to be powered without having to attach the system to the building's internal wiring and the grid.

3. Project Description:

Our system will consist of the Tabuchi Electric EIBS Hybrid Inverter and Storage

Batteries mounted to a portable skeleton and it will be wired for use in pulling from a DC or AC

power source, such as a solar array or the grid. Our system will take a DC input of 70-550V or an

AC input directly off the grid, store the power in two 9.89kWh batteries, and return the power at the standard 120/240V, 60Hz. Via the control panels, power can be inputed or distributed on command, either in on-grid operation or running as a stand alone system. Finally, if setup and installation goes smoothly, we plan to measure capacity, charge and discharge rates, and storage efficiency. However, after reaching the final step of testing the system we found that it had to be plugged into both a DC power source and the AC grid for the batteries and inverter to function.

Due to running out of time we were unable to test the system fully but we were able to confirm that it was wired correctly and that the inverter and batteries turn on.

Figure 3: Tabuchi Diagram of EIBS16GU2 Energy Flow

The project is mobile, with the batteries and inverter mounted separately. With this setup, they can be transported to any accessible site and allowed to charge, then moved to a secondary location for usage. This is accomplished with quick attach-detach cables that interconnect the

separate parts. The important thing to note about the Tabuchi system is that when attached the grid, power will not be allowed to flow back to the grid. If power were able to flow back to the grid then we would have to get certified technicians to set the system up for us, but luckily this is no problem and we can hook to the system to the grid without having to worry about sending power back.

4. Market Research:

The Tabuchi system available for use is approximately 20 kWh in energy capacity and includes an inverter, battery connection box, remote controller, and a ten year warranty. The invoice accompanying this product lists the sale price to Cal Poly as $12,000. The Tesla

Powerwall is the direct competition of Tabuchi Electric’s EIBS, with a prominent presence in

North America. Looking at the current iteration of Tesla’s website, a single powerwall unit has a capacity of 13.5 kWh, includes an inverter, has a 10 year warranty, and can be expanded using multiple connected units. Tesla only sells the powerwall as a set of two for $13,400 with $1,100 of supporting hardware required for a total cost of $14,500. Costs of $12,000/20 kWh and

$7,200/13.5 kWh for Tabuchi EIBS16GU2 and Tesla Powerwall, respectively, equate to $600/1 kWh for Tabuchi and $533.33/1 kWh for Tesla. In the North American market there is no competition in name recognition between the two, other companies with name recognition like

LG are releasing similar systems in 2018/2019. Tabuchi might have less name recognition with an everyday consumer such as a resident, but for the purposes of this project the EIBS will provide useful information on how these systems operate. We learned later in the project that

Tabuchi USA was ceasing operations but the company still works and we were able to contact them with any questions we had while wiring and powering up the system.

5. Customer Archetype:

This P.V. hybrid inverter system that we are assembling has many competitors that have similar products making the market a lot harder to enter. However, our idea of making it a mobile power system is an idea that has not previously been marketed by other companies. The customers that may be interested in this include: Residential Housing, Small Businesses, and

Educational Facilities.

Table 1: Customer Archetype Description Reason Product Use Residential Housing This will be a large Due to the falling cost Our system will be used archetype target. Solar of alternative energy by the average person is growing in popularity solar is more popular at their house. The solar and many people are than ever and with the panels will save them investing. additional backup money on electricity battery system our while the battery option may be even backup will keep more enticing to the essential things average person. powered in the case of a power outage. Small Businesses These are businesses By having a battery The system will power that don’t have a lot of storage system that will the small business at infrastructure but have give power in the event any time the power essential systems that of a power outage, goes out. Saving them need to be on for their small businesses will be money by charging business to operate. able to continue with solar and keeping operations and make the business operational money instead of being when other competitors forced to close and may have to close. missing out on potential revenue. Educational Facilities School campuses need Schools have many The solar will charge power to operate on a tools to educate the batteries without day to day basis. In this students effectively. having to tend to them, day and age technology When these systems go so when the power goes is often needed to teach down unexpectedly out all of the effectively. often times the students computers, projectors, will not actually learn and internet will stay everything that was active and available so planned for the day and the students can learn fall behind. effectively.

It is important to note that after testing the system the end user will have to have the system hooked to both a DC source and an AC source for the system to function properly. Further testing may reveal that once set up we will be able to move the system and discharge the batteries somewhere else.

6. Market Description:

The energy storage market is beginning to emerge in parallel with the economizing of residential-scale solar arrays and the increasing portability of battery-based storage. Companies like Tesla are thrusting themselves into to the public lens of progress. Energy storage technology has thus spent its fair share of time in the recent limelight. In the wake of Hurricane Maria, Elon

Musk and Tesla associates oversaw installation of backup power reserves in Puerto Rico. This is one of the first large-scale implementations of residential-level battery storage, and demonstrated the utility of these systems when residents equipped with Tesla Powerwalls were able to retain power through a recent island-wide blackout according to Electrek.com. At Cal Poly there is also large PV panels on 18.5 acre piece of land west of the Cal Poly campus. The PV panels on this land generate 4.5 megawatts of power which is 25% of Cal Poly’s needs. This PV system also has environmental benefits and has saved Cal Poly about $10 billion on utility bills over the last

20 years. It is important to also mention that in 2020 every new home that is built in the state of

California will be required to be built with solar panels, making our system more viable than ever before.

From this and other examples, we can glean how energy storage will eke its way into the market in the coming years. As it stands, an ever increasing number of people are looking to add storage capabilities to their homes. This is caused by a multitude of trends; the willingness to reduce carbon footprints, the desire to reduce energy expenditure costs, and for sustainability in case of disaster. As an emerging field of technology, this can be capitalized as such. A feasible and very practical business model can be formed around the energy storage market.

In new studies, methods of utilizing quick-response energy storage are being developed and economized. Until recently, the only viable and established way to sell back energy to the grid was to gather and immediately sell energy during the day, if not being used. Similarly, as large-scale (ie, city-wide supply) renewable energy sources become ubiquitous, so too do the problems they introduce. Solar and wind power in particular are vulnerable to inclement weather.

If a given day has lower than expected wind velocity, or is unexpectedly cloudy, the power for an entire region may be compromised. It is in this regard that the nature of the daily energy producing, buying, and selling market is changing; unfettered access to surplus energy in the form of distributed storage has proven to be a solution to a power grid that has low inertia.

The market for commercial and residential energy storage is large, and rapidly growing.

According to Mordor Intelligence, in 2017 about 1,315 energy storage projects were operational with a global installed energy storage of around 176 GW. This figure has continued to climb at an astounding rate in the US, with a 46% year-on-year growth rate for energy storage systems from 2016 to 2017. This means that demand is high for our product and others like it.

7. Business Model Canvas Graphic:

Figure 4: Business Model Canvas

8. Marketing Requirements:

One benefit in the development of energy storage is that is is almost entirely modular, and near infinitely scalable due to the ability to continue adding batteries for additional storage.

The same principles apply at either a residential, industrial, or national level. Each stage requires satisfaction of the same standards: interaction with the power grid, interaction with other electrical products, and interaction with a Photovoltaic system. With this mindset in place, we can begin to start small with our model, a portable and interchangeable merging of inverter and batteries. Designing a simple, residential-scale, proof of concept product will allow us to expand into either a market of larger scale technology, or wider distribution.

This begs two of the most critical questions when engineering a product- how big is it?

And how much will it cost? Luckily for us, here at Cal Poly we have a large surplus of unused

PV supplies on-site. Each Sunpower 435 panel is four feet wide, by six feet long, and ideally we could gain access to these resources at a low cost. Access to these supplies would bolster the resilience of the system, in addition to reducing up-front developmental costs. To start, we will try to spec an approximate size of the panel array, and compare that to the time it would take to charge the Tabuchi energy storage system using N number of panels.

Table 2: Number of Panels in Configuration and Relevant Metrics

Number of Panels Output Voltage (V) Output Current (A) Power (W) Charge Time (Hrs) 2 72.9 10 729 27.43484225 4 145.8 10 1458 13.71742112 6 218.7 10 2187 9.144947417 8 291.6 10 2916 6.858710562 10 364.5 10 3645 5.48696845

Figure 5: Graphical Representation of the Number of Panels in PV Array vs Charge Time

Minimum viability is reached around four panels if we want to meet the full charge cycle per day threshold. Ideally, six or eight panels should be implemented into this project, to account for inclement weather, and allow for rapid power cycling of the system.

9. Block Diagram:

Figure 6: Block Diagram of How the Tabuchi System will be Implemented Our project centers on the Tabuchi Hybrid Inverter (Seen center). This can be input either AC grid power, or DC power from a PV array. From the inverter, the power will be transferred to the storage batteries, mounted separately for ease of mobility. The batteries can then be transported to wherever the power is needed.

10. Requirements: Table 3: Requirements

Requirements Method Justification

Portability The Inverter and Batteries will be Mobility will allow power to be delivered to mounted on seperate mobile rigs wherever it is most needed

Charge/Discharge Rate An array of 8-12 PV panels will The system must be able to charge and allow for a charge/discharge discharge in a day to be useful in future cycle in the course of one day applications

Flexibility The system will be able to draw With a choice of sources, the system will power from either grid or solar provide a variety of testing options sources

Capacity Two Lithium-ion batteries will High capacity storage is needed in order to provide 19.96kWh of storage perform future tests with the system

11. Cost Analysis: Table 4: Bill of Materials

12. Team Coordination:

13. Design Iterations and SolidWorks

Our first thoughts on designing a cart can be found in appendix B of this report. In this appendix we mention that we wanted to design two separate carts, one for the batteries and one for the inverter, due to better mobility and cheaper cost. However, looking further into this design we decided to resort back to the single cart design because the design would be cheaper and the batteries and inverter could remain hooked together making it easier to use. The first designs of this cart were constructed using 8020 as the building material. The designs for the first iteration cart are shown in Figure 7 and Figure 8. The two batteries are located in the front of the cart while the inverter is attached to the rear of the cart. The second layer of the cart is used to keep the batteries and the inverter stationary while the cart is in motion.

Figure; 7: First Iteration of 8020 Cart in SolidWorks

Figure 8: 2D CAD View of First 8020 Cart Design

After looking at our available budget to build the first iteration cart made of 8020 we decided that it would be too costly and looked to different building material options. After talking with

Professor Alhgren about what materials might be useful he suggested Unistrut. Unistrut is a similar like beam construction material like 8020 but it is made of steel instead of aluminum.

Looking at the cost of Unistrut it seemed like it would be within our budget to build the cart with

Unistrut but in order to make the cart lighter we would have to use the thinner version of the

Unistrut. To ensure that the thinner beams could withstand the heavy batteries and inverter, synthetic stress tests were performed on the material in SolidWorks. These stress tests showed us that the thinner Unistrut beams would in fact be able to withstand the weight of the batteries. In light of this we decided to rebuild our cart design in SolidWorks using the Unistrut material. The second iteration cart design made of Unistrut is shown below in Figure 9. It has the same general design as the first iteration but utilizes the new material.

Figure 9: 2nd Iteration of Cart with Unistrut

Our first concern when designing the cart with the Unistrut was that we feared it would not work out as nicely as the 8020. The Unistrut only has two faces in which fasteners can be used to attach it while all four faces of the 8020 could be used to fasten to. However, we were able to make a design that required no drilling into the side of the Unistrut which alleviated these concerns. Our next concern was how easy it was going to be to access the inverter if need be. To fix this problem, we considered how the inverter was designed to be mounted to a wall, so to fit it onto the cart and make it accessible we needed to mount it to our own pseudo wall on the cart.

The third and final iteration of our cart design is shown below in Figure 10.

Figure 10: Final Cart Design

All of our cart designs were created using a program called SolidWorks that is readily used in industry and is also available for free to Cal Poly students who can download the software through the Technical Service Request tab of the Cal Poly Portal. The first step when designing the carts was going to the 8020 and Unistrut websites where 3D CAD files can be found for each part that we desired. All of the part files were found relatively easily in the technical section of each part. Each part was then loaded into SolidWorks and cut down to the appropriate size that we needed in each design. Next, an assembly had to be built in SolidWorks.

An assembly in SolidWorks is a file that consists of different part files that have been assembled and mated together in a certain way. It is also important to note that when opening an assembly file, the computer used must have all the individual part files downloaded as well. Once the assembly file had been created, the part files that were modified to be the correct lengths are then added to the assembly. To join the parts in the correct orientation that is desired the mate tool is used. The mate tool connects faces and edges of adjacent parts together so that they may become fully defined in the assembly. This process of loading in parts and mating them is then repeated until the final design is created.

For the materials we used we also used the stress testing tool in SolidWorks to ensure the material we were using could withstand the weight of the batteries and the inverter. This is done by loading a part into SolidWorks and choosing points to fix in space. These points will be on the end of the part in our case because this is where the Unistrut beams will be attached to the rest of the cart. Finally, The stress must be added to the system and for our project the stress of the batteries would be in the middle of the beams. The stress that was added was equivalent to

the weight of the batteries plus a little bit more just in case somebody wants to sit on the cart or something like that. The stress test showed us that the material may flex a little bit with the weight of the battery but would ultimately be able to withstand the force.

Table 5: Budget for Unistrut Cart Design

Cost Part per Total Item Name Number Manufacturer Description Unit # Units Cost Purchase Link

Unistrut Based Cart

HHXN - Hex HHXN0 Nut 50 Unistrut Hex Nut 150 https://www.unistrut.us/product-details/hhxn

P1026 (Cheape P1026 - 2 st Hole, 90 Material 2 Hole "L" degree Fitting ) Unistrut Bracket 20 https://www.unistrut.us/product-details/p1026

P3045 - 2 Hole, "Z" 2 Hole "Z" Shape Fitting P3045 Unistrut Bracket 18 https://www.unistrut.us/product-details/p3045

P1036 - 3 Hole, Flat Plate 3 Hole, Flat Fitting P1036 Unistrut Plat Fitting 16 https://www.unistrut.us/product-details/p1036

P3300T P3300T - (Cheape 1-5/8" X 1-5/8" X 7/8", st 7/8", 12 12 Gage, Material Gage, Slotted ) Unistrut Slotted (20') 4 https://www.unistrut.us/product-details/p3300t

HHCS0 50094 HHCS - Hex 15/16E Hex Head Head Screw G Unistrut Screw 150 https://www.unistrut.us/product-details/hhcs

P3006 Channel Inter-channe https://www.unistrut.us/product-details/p3006-thr Nut P3006 Unistrut l bracer 125 u-p3013

https://www.unistrut.us/product-details/p1062-thr Spacer Bracket P1062 Unistrut .25" Spacer 2 u-p2490

Table 6: Budget for 8020 Cart Design 8020 Based Cart

Aluminum 2.00” X 2.00” T-Slotted Profile - Structure https://8020.net/2 Eight Open T-Slots 2020 80/20 Inc. (8X 102") $0.57 816 $465.12 020.html

Aluminum 2.00” X 2.00” T-Slotted Profile - Structure https://8020.net/2 Eight Open T-Slots 2020 80/20 Inc. (1X 38") $0.57 38 $21.66 020.html

8 Hole Inside 10 Series 8 Hole - Inside Corner Corner https://8020.net/4 Bracket 4114 80/20 Inc. Bracket $5.35 30 $160.50 114.html

4 Hole Inside 10 Series 4 Hole - Wide Inside Corner https://8020.net/4 Corner Bracket 4113 80/20 Inc. Bracket $4.05 34 $137.70 113.html

Bolt Assembly Bolt Assembly: 1/4-20 x .500" Black for Corner https://8020.net/3 BHSCS 3393 80/20 Inc. Bracket $0.40 504 $201.60 393.html

Caster 10 Series Flange Mount Caster Mount https://8020.net/2 Base Plate 2418 80/20 Inc. Base Plate $15.75 4 $63.00 418.html

Bolt Assembly for Caster Bolt Assembly: 1/4-20 x .500" Black Mount https://8020.net/3 SHCS 3491 80/20 Inc. Base Plate $0.37 16 $5.92 491.html

Plate for Attaching 10 Series 12 Hole - 90 Degree Corners of https://8020.net/4 Angled Flat Plate 4128 80/20 Inc. Cart $10.80 16 $172.80 128.html

Estimated Subtotal $1,228.30

Table 7: Budget for Electrical Components Cables and Additional Connections

8 gage cabling used to connec https://www.homedepot.com/p/Southwir 2048830 t the e-100-ft-8-Black-Stranded-CU-SIMpull- 8 AWG Wire, 100ft 2 Southwire inverter $45.97 2 Spools $91.94 THHN-Wire-20488302/202519272

Connec tions for 8 https://www.homedepot.com/p/Gardner- 8 AWG Ring Gardner AWG Bender-8-AWG-1-4-in-Tab-Ring-Termin Terminal 15-095 Bender wire $3.34 3 $10.02 al-Vinyl-Red-5-Pack-15-095/205846740

Discon nect box prevent ing https://www.homedepot.com/p/GE-100- inverter Amp-240-Volt-Fusible-Outdoor-General 100 Amp 240-Volt damag -Duty-Safety-Switch-TG3223R/2029786 Fusible Disconnect TG3223R GE e $96.11 1 $96.11 56

Protect s the inverter from https://www.homedepot.com/p/Cooper- current Bussmann-20-Amp-250-Volt-Fusetron- 20 Amp 250 Volt FRN-R-2 Cooper over HD-Time-Delay-Cartridge-Fuse-FRN-R- Fuzetron Fuse 0 Bussmann 20A $5.47 2 $10.94 20/100188109

WIRING SUBTOTAL 209.01

14. Assembly Iteration and Challenges We began our construction with four 20 foot beams of 12 Steel Unistrut which we then cut down to size to match our CAD design. Each piece was measured and cut with a miter saw rented from the Cal Poly machine shop.

From there, each piece was transported back to building 20, where they were laid out and categorized for assembly.

Assembly began with the baseplate ring, which would serve as the bottom supporting structure for the remainder of the cart. Utilizing ½ inch bolts as well as a series of adapter plates and spring-locking channel nuts the unistrut could be easily fitted together. From this base, uprights were constructed to hold the second layer and caster wheels were fitted to allow for our desired mobility. However, upon completion of the second layer we met our first obstacle. The heads of the bolts inserted into the channels of the unistrut were too bulky to be reached by a ratchet or wrench. As such, these bolts were only able to be tightened minimally with pliers. Being unsatisfied with the unstable result, we set about performing an almost entire rebuild. This required us to drill through some parts of the Unistrut to enable the use of bolts that would travel through the entire structure and allow us to put more torque on the bolts when

tightening them. Moving past this issue, we mounted the batteries, and secured them via the second layer’s ring. We began to mount the inverter as well, but again ran into a snag. We had previously thought that the two upright bars would be strong enough to support the inverter and prevent tipping, but we felt as though this design may not hold up under repetitive relocation. Using L and S brackets, we affixed the upright bars to our second layer, causing the cart to be much more stable overall.

Next, we mounted the DC disconnect and connection box to the rear of the inverter, utilizing open space as efficiently as possible. This was a challenge, as the devices had differing space requirements and wiring methods to consider. Once all of the peripherals were mounted to the cart, we started on the wiring. We cut our measured wire to size, with 25 feet being dedicated to each of the three grid-tied connections. The remainder of the wire was used to interconnect the system. We ran the 25 foot length wire first, using Hampden plug heads on one end and connecting the other to our DC disconnect.

The disconnect (With three 20 Amp fuses installed) was then wired to the 240V split phase input of inverter by way of crimped ring terminals for easy connection/disconnection.

After this, we completed the interconnection by wiring the batteries to the inverter via the

connector box using 10 gauge wire. Finally, we interconnected the batteries, inverter, and two current monitors to the control panel, allowing the entire system to be controlled simultaneously.

15. Project Significance and Applications This project is centered around a battery-based storage system or BESS. This BESS is designed for a consumer as a DC-coupled solar energy storage system, which consists of two 10 kWh, 86 V batteries, a grid-tie or standalone inverter, and some control electronics. These nature of these control electronics allows for this group to manipulate the operation of the BESS to suit the needs of this Senior Project.

The BESS is capable of operating on its own as an islanded system and operating as a grid-tied DC to 240 V split-phase system. Both of these generalized operational modes are useful to this project and the ability to island is essential to the predicted uses of this system. There are more details to these operational modes, but it is important to note that the BESS is capable of charging on AC and DC, and discharging via AC.

The Tabuchi BESS is capable of operating strictly as a standalone system, however this group wishes to use the system in conjunction with the Laboratory Voltage Distribution System

(LVDS) for charging purposes. The BESS is potentially capable of absorbing, storing, and supplying power to the LVDS. As this project stands right now, the drawing of supplemental power from the BESS to the LVDS is not feasible. It is possible however to charge the BESS because the LVDS is capable of supplying the 240 V split-phase AC power that the BESS is designed to handle when operating in grid-tied mode. Once charged, the BESS would need to be disconnected from the LVDS and operated in standalone mode in order to discharge the stored energy through the user’s choice of load. The transition between modes of operation is achieved

via the control panel of the BESS and the disconnection of the BESS from the LVDS is achieved in 20-102 by disconnecting the appropriate plugs, a similar activity is done regularly by students in the EE 295 and EE 444 labs.

The LVDS is an older system that was not designed with storage in mind. The Cal Poly

Outdoor Solar Lab, an addition planned for the near future outside of building 20 of the Cal Poly campus, could feature infrastructure which could better incorporate the BESS as a storage device capable of supplying supplemental power. These features would need to be discussed with the appropriate authorities.

16. Peripherals The main peripheral necessary at the time of assembly for this project was the addition of an AC disconnect box with 20 amp fuses inside. This was added in place of what would have been the inverter’s connection to the residential electrical panel. The disconnect was deemed necessary in order to assure that the system could be safely disconnected from the source or load; the fuses inside the disconnect are to assure that if a fault were to occur within the system then the system’s components as well as the connected load or charging system would incur minimal damage.

To make the AC connections for charging/discharging the system a set of 8 AWG stranded copper conductors were fitted with Hampden Engineering Co. plug connectors. These plugs allow the project to connect directly to the Hampden sockets on the wall of Cal Poly

Building 20 Room 102, giving the system access to 240 V split-phase AC power.

There are a few options for expansion of this project with the addition of further peripherals. These peripherals were either inaccessible at the time of this project’s assembly or were outside of the budget available to the project team at the time of this project’s assembly.

The first not-yet-realized peripheral is the option to charge via solar panels which would require a DC connection to be wired into the inverter of the BESS. This is the hope of those professors who oversaw this project: that the BESS might be integrated into a solar-inclusive microgrid of some kind on the Cal Poly campus. Our team has circulated talk regarding a rooftop installation of an array of panels, to be fed into room 102, and take up space on the power wall.

This allows for truly renewable, standalone DC power supplied to the BESS by simply maneuvering the cart into room 102 and connecting. Ben Johnson and other maintenance personnel at Cal Poly

Another set of peripheral options is the conversion of the system’s AC connection from

240 V split-phase to 208 V three-phase power. The desire for this conversion was expressed by professor Ahlgren regarding the creation of a three-phase microgrid within Engineering East on the Cal Poly campus. This conversion would require two devices: a phase converter and a transformer. It is the belief of this group that the cheaper of the two possible setups for this conversion would be to transform 240 V single-phase to 208 V single-phase and then convert single-phase to three-phase. This addition to the system would be expensive but could allow for more learning opportunities regarding three-phase power which is still very relevant in today’s power systems.

Figure 11 : Phase Converter to be Purchased in Future

17. Project Continuation

So far in this project we have completed the task of making the system mobile and power the system when tied to the “grid” when plugged into the 240 V split phase power coming from the wall in room 102. We have also managed to do all the wiring that connects the batteries to the inverter.

From this point moving forward, the biggest step that must be taken is either plugging the system into a PV array or equip more wires with Hampden plug heads so that we can interface the system into DC power from room 102 in order to bypass the need for a PV system. This must be done to set up the inverter and fully turn it on and have access to the batteries. When we powered the system on when it was only connected to the grid the system would not allow us to

configure the converter or access the charge that was stored in the batteries. Once the batteries power can be accessed, what we do with that power would be up to the user. Our thought was that we should have a converter that can take the DC power that is output from the batteries and convert it into three phase power that could potentially be used in a EE lab in the future. Other options may be viable but this is the one that came to our minds.

One last change to our design that may be considered as well is the wheels that make the cart mobile. The only wheels that we were able to find that somewhat work have a single bolt that holds them in place. Due to the heavy weight of the batteries and the inverter, the wheels can have the tendency to lean as they only have a square washer to push against. If possible, heavy duty caster wheels with a standard four bolt plate might make a better alternative if possible to attach to the Unistrut.

As part of the future work for this project, studying the operation and the model of the

PV-BESS system is of great importance. Three major areas can be divided into studying its control systems, inverters, and how the battery energy systems work together. High penetration of renewable energies in power systems brings opportunities such as zero emission but it can bring challenges too. The alternative nature of renewable energies unlike the conventional sources are keep changing which requires new controlling algorithms for having a stable power system. The first step before proposing new controlling algorithms, is to make sure that the comprehensive modeling of these systems are available (PVBESS in this case). Many efforts have been done in the literature in finding the detailed modeling of microgrids including the inverters [1] that can be used as a guidance for modeling of this system. New proposed controlling algorithm can now be applied to these models for evaluating their performance

before applying on the real-world systems. However, since the number of power electronics components are increasing in the power system, the nature of the controlling methods are changing from the centralized methods to the one that does not rely much on just a single central node. These algorithms can be used among different microgrids to control the power systems’ variables such as the voltage in the distributed mode [2]. In the future, in an effort to connect this

Microgrid system to other parts of the Microgrid at CalPoly, we can investigate these methods and investigate their performance on controlling the voltages and frequencies.

Another important part of the future work that can be considered to be done is to analyze the cost benefit of this system. Since this PVBESS is connected to an alternative source

(photovoltaics system), only gathering an enormous amount of information will enable us to do such an analysis. These data such as the temperature, humidity, cloud direction and speed, time of the day can be collected on a long term basis and then by using big data analytics given in [3], we can find a model that can give us the desired cost benefit analysis.

18. Department Future

As mentioned before, the hope of this project is to be expanded into the to-be-built Cal

Poly Solar lab. This allows for standalone testing and off-grid analysis of power systems.

Eventually, we hope the Tabuchi EIBS can be used as a storage platform in a class session that focuses on renewable energy. Of course, this means design and implementation of a PV array at the Cal Poly EE building is necessary. There are numerous administrative hurdles that need to be tackled prior to the realization of the system’s full potential; large-scale solar projects have been attempted in the past, but have faced strong resistance from Cal Poly at large.

Expanding the power systems subdepartment has the potential to breathe new life into a somewhat dying class of Electrical Engineering. According to former IEEE Power and Energy

Society president John McDonald, the average age for people working in the power sector of

Electrical Engineering is ten years older than the average age of every other IEEE discipline. It is widely considered to be a relatively stagnant field of study; infrastructure is usually old, underfunded, and not as flashy as, say, electronics. However, with the Tobuchi system, we have the platform to show new Cal Poly students that, in fact, there is significant research and modernization occurring in the power sector. This realm of occupation which was historically seen as mundane, has now been recently introduced into the premier club of high tech. With new developments in smart grids, problems involving sustainability, protection, and economization can now be addressed on any scale. The chord which we hope to strike with the audience of

Power Systems is that what has traditionally worked in power is phasing out. On the whole, we need ingenuity and new ideas, and the Cal Poly EE department is the perfect environment to foster that.

19. References: [1] A Dehghan Banadaki, F DoostMohammadi, A Feliachi, State Space Modeling of Inverter Based Microgrids Considering Distributed Secondary Voltage Control”, North American Power Symposium (NAPS). 2017.

[2] AD Banadaki, A Feliachi, V Kulathumani, “Fully Distributed Secondary Voltage Control in Inverter-Based Microgrids”, IEEE Transmission and Distribution Conference (T&D), 2018. [3] A Dehghan Banadaki, T Taufik, A Feliachi, “Big Data Analytics in a Day-Ahead Electricity Price Forecasting Using TensorFlow in Restructured Power Systems” The 2018 International Conference on Computational Science and Computational Intelligence.

[4]“2018 Solar Batteries: Sonnen, Tesla, Aquion, LG Chem | EnergySage,” Solar News, 12-Nov-2018. [Online]. Available: https://news.energysage.com/tesla-powerwall-vs-sonnen-eco-vs-lg-chem/. [Accessed: 08-Dec-2018].

[5]“Energy Storage Market Size, Share - Segmented by Type (Pumped Storage, Lithium Ion, Flow Battery, Lead Acid Battery), By Application (Residential Commercial and Industrial), Location of Deployment, and Geography - Growth, Trends and Forecast (2018 - 2023).,” Analysis of UAE Telecom market segmented by major players, market share, revenues and regulations. [Online]. Available: https://www.mordorintelligence.com/industry-reports/energy-storage-market. [Accessed: 08-Dec-2018].

[6]“EnergySage Releases Its Latest Solar Marketplace Intel Report™ at Bloomberg's Future of Energy Summit,” EnergySage. [Online]. Available: https://www.energysage.com/press/energysage-marketplace-intel-report-6. [Accessed: 08-Dec-2018].

[7]“How Solar Panels Work,” Union of Concerned Scientists. [Online]. Available: https://www.ucsusa.org/clean-energy/renewable-energy/how-solar-panels-work#.XAtOl2hKiiM. [Accessed: 08-Dec-2018].

[8]IISD, “Solar Energy Installations in 2017 Topped Gas, Coal and Nuclear Combined, Amid Steady Growth in Renewables | News | SDG Knowledge Hub | IISD,” SDG Knowledge Hub. [Online]. Available: http://sdg.iisd.org/news/solar-energy-installations-in-2017-topped-gas-coal-and-nuclear-combine d-amid-steady-growth-in-renewables/. [Accessed: 08-Dec-2018].

[9]D. Muoio, “Tesla is sending hundreds of battery packs to Puerto Rico in the wake of major hurricanes,” Business Insider, 30-Sep-2017. [Online]. Available: https://www.businessinsider.com/tesla-battery-packs-puerto-rico-hurricane-maria-2017-9. [Accessed: 08-Dec-2018].

[10]Z. Shahan , “Solar Panel Cost Trends (Tons of Charts),” CleanTechnica, 28-Jul-2016. [Online]. Available: https://cleantechnica.com/2014/09/04/solar-panel-cost-trends-10-charts/. [Accessed: 08-Dec-2018].

[11]“Solar Industry Research Data,” SEIA. [Online]. Available: https://www.seia.org/solar-industry-research-data. [Accessed: 08-Dec-2018].

[12]“Tesla Powerwall,” Tesla, Inc. [Online]. Available: https://www.tesla.com/powerwall. [Accessed: 08-Dec-2018].

[13]J. Weaver, “Residential energy storage grows 9x in Q1 2018,” pv magazine USA, 06-Jun-2018. [Online]. Available: https://pv-magazine-usa.com/2018/06/06/residential-storage-market-grows-1200-quarter-over-qu arter/. [Accessed: 08-Dec-2018].

Appendix A: Senior Project Analysis Project Title: PV Hybrid Inverter BESS

Students: William Dresser, Owen McKenzie, Derek Seamen, Jacob Sussex, Jonathan Wharton

Adviser: Ali Dehghan-Banadaki

1. Summary of Functional Requirements Our project makes the heavy Tabuchi Inverter and batteries that weigh over 700 lbs mobile

and easy to move around to different locations in the EE building. It was designed to fit

through doors and easily move in and out of different rooms. The final product will be able to

be charged by the power in room 102 and then moved to a different location where the power

from the batteries can be used to power labs, machinery, or anything that the user desires. The

system will also be capable of charging from solar panels mounted on the roof of the EE

building.

2.Primary Constraints The biggest challenge that needed to be solved in this project was making the entire Tabuchi

inverter and battery system mobile. This system weighs over 700 lbs and it was difficult

finding a material that would be able to handle this much weight while also staying under our

allocated budget for the project. The cart that was designed to mobilize this system had many

design iterations due to this challenge. We looked at two main materials, 8020 aluminum

beams and Unistrut, to build the cart. Once we decided on the material we had to design a cart

that would allow the end user to access all the components for maintenance or repair in the

future which was also challenging. After settling on a design we had to assemble the cart and

cut all the material down to length. While assembling we had to improvise as we went

because the design that was built in the computer did not work out exactly like we wanted in

reality. Finally, figuring out how to correctly wire the system and do so in a safe manner was

also very challenging. We had to flip through the user manual and use our own knowledge to

complete the wiring correctly in order for the system to operate properly.

3.Economic This system is manufactured by a company that recently went out of business called Tabuchi.

It is a very expensive system as it is complicated to build and required many man hours of

assembly and engineering design to manufacture. In addition to this the cart and the electrical

components that were used to achieve our functionality requirements took time and money for

companies to manufacture so that we could implement them in our design. The costs of this

system during its lifecycle accrue during any maintenance that may be required due to normal

wear and tear. There are also many benefits to this product during its lifecycle such as saving

money using solar cells and battery backup during power outages. Below are the budgets for

the cart and the electrical components that were used in the design.